A milieu of regulatory elements in the epidermal differentiation complex syntenic block: implications for atopic dermatitis and psoriasis

Abstract

Two common inflammatory skin disorders with impaired barrier, atopic dermatitis (AD) and psoriasis, share distinct genetic linkage to the Epidermal Differentiation Complex (EDC) locus on 1q21. The EDC is comprised of tandemly arrayed gene families encoding proteins involved in skin cell differentiation. Discovery of semi-dominant mutations in filaggrin (FLG) associated with AD and a copy number variation within the LCE genes associated with psoriasis provide compelling evidence for the role of EDC genes in the pathogenesis of these diseases. To date, little is known about the potentially complex regulatory landscape within the EDC. Here, we report a computational approach to identify conserved non-coding elements (CNEs) in the EDC queried for regulatory function. Coordinate expression of EDC genes during mouse embryonic skin development and a striking degree of synteny and linearity in the EDC locus across a wide range of mammalian (placental and marsupial) genomes suggests an evolutionary conserved regulatory milieu in the EDC. CNEs identified by comparative genomics exhibit dynamic regulatory activity (enhancer or repressor) in differentiating or proliferating conditions. We further demonstrate epidermal-specific, developmental in vivo enhancer activities (DNaseI and transgenic mouse assays) in CNEs, including one within the psoriasis-associated deletion, LCE3C_LCE3B-del. Together, our multidisciplinary study features a network of regulatory elements coordinating developmental EDC gene expression as an unexplored resource for genetic variants in skin diseases.

Figure 1.

The EDC is comprised of tandemly arrayed gene families. Human (hg18) EDC on chromosome 1q21 (1.6 Mb).

Figure 2.

Coordinated expression of EDC genes during epidermal differentiation and barrier formation. (A) Mouse (mm9) EDC (chr3), 3.1 Mb, are comprised of 4 gene families (FLG-like [II], LCE [III], SPRR [IV] and S100 [I and V]). Group I represents a cluster of 2 S100 genes (S100A10, S100A11). (B) Heatmap reflecting a semiquantitative real-time PCR analysis of EDC gene expression from mouse dorsal skin at E15.5 (epidermal differentiation) and E16.5 (barrier formation). Experiments were done in triplicate (n = 2 per embryonic stage) and normalized to β₂-microglobulin. Gray scale legend, fold-change over E13.5 expression.

Figure 3.

The EDC represents an ultraconserved microsyntenic block in mammals. Each EDC loci from human, chimp, rhesus, mouse, rat, dog and opossum is depicted as locally collinear blocks (LCBs) (colored boxes) representing homologous regions shared by the aligned genomes and does not contain any rearrangements. Vertical lines trace homologous LCBs between genomes. LCBs below a genome's center black line are in reverse complement orientation relative to the human reference EDC locus. Red vertical lines mark interchromosomal boundaries or chromosomal distances <100 000 kb (opossum). +, forward orientation; −, reverse complement. Note that the S100 genes are in reverse complement in the dog.

Figure 4.

Regulatory activities in the CNEs of the EDC. (A) CNEs (grouped into clusters I–V, as depicted in Fig. 1) are labeled as distance (kb) from the S100A10 transcriptional start site. Identification of CNEs span (hg18) chr1:150 202 011–151 891 137 including −20 kb of the transcriptional start site of S100A10 and +20 kb downstream from S100A1 transcript. EDC CNEs were tested for in vitro enhancer and repressor activity (luciferase reporter assays) in keratinocytes under (B) differentiating and (C) proliferating conditions. CNEs exhibiting >2-fold increased luciferase activity demonstrate enhancer activity and those that exhibit >2-fold decreased luciferase activity demonstrate repressor activity. Columns represent an average of two independent experiments performed in duplicate. Numbered CNEs are highlighted in designated boxes where Rectangle = enhancers (differentiating and proliferating), Hexagon = repressors (differentiating and proliferating), Triangle = enhancers (differentiating only), Oval = repressors (proliferating only), Bars, standard error.

Figure 5.

h923 and h621 in LCE3C_LCE3B-del function as enhancers in vivo. Enhancer activity was assayed by DNaseI hypersensitivity and transgenic hsp68-lacZ mice. (A) PCR amplicons of tiling primer sets (–6) targeting h923 are indicated. (B) DNaseI sensitivity (ΔC_T = ΔC_T[DNase-treated] − ΔC_T[No DNase]) of individual primary keratinocyte samples (#1, #2) of h923. Light grey = low DNaseI, dark grey = hi DNaseI. Results are an average of triplicate readings. Bars, standard error. (C) h923-hsp68-lacZ transgenic mice (F0) demonstrate β-galactosidase-staining in the skin. All β-galactosidase-stained h923-hsp68-lacZ mice (two out of five genotyped-lacZ+ F0 mice) stained skin-specific at E16.5 (shown) and at pnd0. (D) h621-hsp68-lacZ (F0) demonstrate β-galactosidase-staining in the skin at E16.5. Five h621-hsp68-lacZ mice genotyped positive for lacZ. Four were β-galactosidase positive and two were skin-specific (shown). The remaining two β-galactosidase positive mice showed non-specific staining in the paws and brain (data not shown). H&E stains of β-galactosidase-stained cross-sections of the skin, Dotted lines indicate basement membrane demarcating the epidermis from the dermis, 40×.